Back to rankings

Lovable Cloud: Pause Projects and Resize Instances from Chat

Lovable

Lovable gave users direct, chat-driven control over their Lovable Cloud backend resources. Users can now manually pause a project's Cloud backend to stop compute-credit consumption when not actively working on it, and can ask Lovable in natural language to resize compute instances without leaving the chat interface. Both changes target cost control and performance tuning for teams running production or staging backends on Lovable Cloud.

Pausing Cloud Backends on Demand

Lovable introduced a manual pause control for Lovable Cloud projects, letting users stop a backend from consuming compute credits during periods of inactivity. The control lives under Cloud tab → Overview → Advanced settings, where selecting Pause and confirming immediately suspends the project's database, authentication, storage, and edge functions. While paused, the live app stops functioning until the project is resumed, making this best suited for staging environments, side projects, or apps between active development sprints rather than production apps serving live traffic.

This addresses a common pain point for teams juggling many Lovable Cloud projects: credits previously kept accruing on idle backends with no easy way to halt spend short of deleting the project outright.

Resizing Compute Instances Through Chat

Alongside the pause control, Lovable added the ability to resize a project's Cloud compute instance directly from the chat interface, instead of requiring a trip to Advanced settings. Users can describe a performance problem in plain language, such as a backend that slows down under load, and Lovable responds with a size picker ranging from Tiny to Large, with the project's current size preselected.

Why It Matters

Together, these two features push Lovable Cloud further toward being a genuinely operable backend platform rather than a black box. Pausing gives cost-conscious builders a lever to control spend without losing project state, and chat-driven resizing lowers the barrier to diagnosing and fixing performance issues for users who are not comfortable navigating infrastructure settings menus.